Professional Documents
Culture Documents
Assistant
Adrian Egger, HIL E13.1, email: egger@ibk.baug.ethz.ch
Course Website
Lecture Notes and Homeworks will be posted at:
http://www.chatzi.ibk.ethz.ch/education/method-of-finite-elements-ii.html
Suggested Reading
“Nonlinear Finite Elements for Continua and Structures”, by T.
Belytschko, W. K. Liu, and B. Moran, John Wiley and Sons, 2000
“The Finite Element Method: Linear Static and Dynamic Finite
Element Analysis”, by T. J. R. Hughes, Dover Publications, 2000
Lecture Notes by Carlos A. Felippa
Nonlinear Finite Element Methods (ASEN 6107): http://www.
colorado.edu/engineering/CAS/courses.d/NFEM.d/Home.html
Institute of Structural Engineering Method of Finite Elements II 2
Course Outline
Eigenvalue Problems
Special Topics
- The Scaled Boundary Element & Extended Finite Element methods
Example
The Axially Loaded Bar
Continuous Discrete
q|y+dy
q|x
dy
q|x+dx
dx
q|y
h1
h2
Flow
L of water
Permeable Soil
Impermeable Rock
∂2φ ∂2φ F = KX
k ∂2x + ∂2y =0
Laplace Equation Direct Stiffness Method
Continuous Discrete
q|y+dy
q|x
dy
q|x+dx
dx
q|y
h1
h2
Flow
L of water
Permeable Soil
Impermeable Rock
∂2φ ∂2φ F = KX
k ∂2x + ∂2y =0
Laplace Equation Direct Stiffness Method
d du
c(x) + f(x) = 0
dx dx
d
− c(0) u(0) = C1 (Neumann BC)
dx
u(L) = 0 (Dirichlet BC)
Constitutive
Physical Problem (1D) Diff. Equation Quantities
Law
T=temperature
A=area Fourier
One dimensional Heat 𝑑 𝑑𝑇
�𝐴𝑘 � + 𝑄 = 0 k=thermal 𝑞 = −𝑘 𝑑𝑇/𝑑𝑥
flow 𝑑𝑥 𝑑𝑥 conductivity 𝑞 = heat flux
Q=heat supply
u=displacement
A=area Hooke
𝑑 𝑑𝑢
Axially Loaded Bar �𝐴𝐸 � + 𝑏 = 0 E=Young’s 𝜎 = 𝐸𝑑𝑢/𝑑𝑥
𝑑𝑥 𝑑𝑥 modulus 𝜎 = stress
B=axial loading
Z Z Z X
Wint = ¯T τ dΩ = Wext = ūT bdΩ + ūST TS dΓ + ūiT RC i
Ω Ω Γ i
where
TS : surface traction (along boundary Γ)
b: body force per unit area
RC : nodal loads
ū: virtual displacement
¯: virtual strain
τ : stresses
Z l Z l
0
wg dx = [wg ]l0 − gw 0 dx
0 0
Z l Z l
0 0
w cu dx = wfdx + w (0)C1
0 0
S = {u|u ∈ C 0 , u(l) = 0}
S 0 = {w |w ∈ C 0 , w (l) = 0}
Institute of Structural Engineering Method of Finite Elements II 15
Weak Form
Notes:
𝑥1𝑒 𝑥2𝑒
Then break the overall integral into a summation over the finite
elements:
" #
X Z x2e Z xe
2
w 0 cu 0 dx − wfdx − w (0)C1 = 0
e x1e x1e
Galerkin’s method assumes that the approximate (or trial) solution, u, can
be expressed as a linear combination of the nodal point displacements ui ,
where i refers to the corresponding node number.
X
u(x) ≈ u h (x) = Ni (x)ui = N(x)u
i
where bold notation signifies a vector and Ni (x) are the shape functions.
In fact, the shape function can be any mathematical formula that helps us
interpolate what happens at points that lie within the nodes of the mesh.
In the 1-D case that we are using as a reference, Ni (x) are defined as 1st
degree polynomials indicating a linear interpolation.
!
Z l X dNi (x) dN j (x)X
ui c wj dx
0 dx dx
i j
Z l X X
− f wj Nj (x)dx − wj Nj (x)C1 =0
0 j j
x=0
"Z ! #
l
X X dNi (x) dNj (x)
wj cui − fNj (x)dx + (Nj (x)C1 )|x=0 = 0
0 dx dx
j i
The above equation has to hold ∀wj since the weighting function w (x) is
an arbitrary one. Therefore the following system of equations has to hold:
Z l X !
dNi (x) dNj (x)
cui − fNj (x)dx + (Nj (x)C1 )|x=0 = 0 j = 1, ..., n
0 dx dx
i
After reorganizing and moving the summation outside the integral, this
becomes:
" #
X Z l dNi (x) dNj (x) Z l
c ui = fNj (x)dx + (Nj (x)C1 )|x=0 = 0 j = 1, ..., n
0 dx dx 0
i
Ku = f
where writing the integral from 0 to l as a summation over the
subelements we obtain:
Z x2e Z x2e
e e
K = Ae K −→ K = NT
,x cN,x dx = BT cBdx
x1e x1e
Z x2e
f = Ae f e −→ f e = NT fdx + NT h|x=0
x1e
𝑥 𝜉
𝑥1𝑒 𝑥2𝑒 −1 1
d 1 1
−1 1
where N,ξ = 2
(1 − ξ) 2
(1 + ξ) = 2 2
dξ
dx x e − x1e h
and x,ξ = = 2 = = J (Jacobian) and h is the element length
dξ 2 2
dξ
ξ,x = = J −1 = 2/h
dx
From all the above,
c 1 −1
Ke =
x2e − x1e −1 1
Similary, we obtain the element load vector:
Z x2e Z 1
fe = NT fdx + NT h|x=0 = NT (ξ)fx,ξ dξ + NT (x)h|x=0
x1e −1
Exterior
model
95% are
shell
elements
Internal structure
zoom. Some Brick
FEM model: and tetrahedral
150000 Nodes elements
http://www.colorado.edu/engineering/CAS/Felippa.d/FelippaHome.d/Home.html
υ´ υ´
υ
Equilibrium
– shear force
(1)
Free end with applied load
(2)
(S) (3)
Simple support
(4)
Clamped support
(5)
The order of the shape functions is chosen based on the problem’s strong
form. Moreover, the shape functions are usually expressed in the
iso-parametric coordinate system
Bar Element
d 2u
For the standard bar element, the strong form is AE + f (x) = 0. The
dx 2
d2u
homogeneous form is AE dx 2 = 0 which is a 2nd order ODE with known solution:
u(x) = C1 x + C2 , i.e., a 1st degree polynomial.
The shape functions N(ξ) are therefore selected as 1st degree polynomials:
1
(1 − ξ)
→ u1
N(ξ) = 2
1
→ u2
(1 + ξ)
2
Beam Element
d2 d 2υ
For the standard beam element, the strong form is EI 2 + f (x) = 0.
dx 2 dx
2 2
d d υ
The homogeneous form is EI 2 = 0 which is a 4th order ODE with
dx 2 dx
known solution: u(x) = C1 x3 + C2 x 2 + C3 x + C4 , i.e., a 3rd degree polynomial.
The shape functions H(ξ) are therefore selected as 3rd degree polynomials:
1
(1 − ξ)2 (2 + ξ)
4 → u1
1
(1 − ξ)2 (1 + ξ) → θ1
4
H(ξ) =
1
(1 + ξ) (2 + ξ) → u2
2
4
1 → θ2
2
(1 + ξ) (ξ − 1)
4
Institute of Structural Engineering Method of Finite Elements II 34
Cheat Sheet
Basic Formulas-Stiffness Matrices
Going from Strong to Weak form yields the expressions of the equivalent
stiffness matrix for each element:
Bar Element
Z L
dNj (x) dNi (x)
Kij = AE dx
0 dx dx
Going from Strong to Weak form yields the expressions of the equivalent
stiffness matrix for each element:
Beam Element
L
d 2 Hj (x) d 2 Hi (x)
Z
Kij = EI dx
0 dx 2 dx 2
Governing Equations
Equilibrium Eq: ∇s σ + b = 0 ∈Ω
Kinematic Eq: = ∇s u ∈Ω
Constitutive Eq: σ =D · ∈Ω
Traction B.C.: τ · n = Ts ∈ Γt
Displacement B.C: u = uΓ ∈ Γu
Divide the body into finite elements connected to each other through
nodes
Iso-parametric Mapping
4
X
x= Ni (ξ, η)xie
i=1
X4
y= Ni (ξ, η)yie
i=1
Z Z Z
Ktot · d = f tot where Ke = BT DBdΩ, fe = N T BdΩ + N T ts dΓ
Ωe Ωe ΓeT
*The total stiffness Ktot is obtained via assembly of the element matrices Ke .
The same holds for the force vectors.
Institute of Structural Engineering Method of Finite Elements II 40
Axially Loaded Bar Example
Assumptions:
The cross-section of the bar does not change after loading.
The material is linear elastic, isotropic, and homogeneous.
The load is centric.
End-effects are not of interest to us.
du d 2u
Also, = , σ = E , q(x) = ax ⇒ AE 2 + ax = 0
dx dx
Strong Form
d 2u
AE + ax = 0
dx 2
u(0) = 0 essential BC
du
f(L) = R ⇒ AE = R natural BC
dx x=L
Analytical Solution
ax 3
u(x) = uhom + up ⇒ u(x) = C1 x + C2 −
6AE
C1 , C2 are determined from the BC
Institute of Structural Engineering Method of Finite Elements II 44
Axially Loaded Bar Example
An analytical solution cannot always be found
Approximate Solution - The Galerkin Approach (#3): Multiply by the weight function
w and integrate over the domain
Z L d 2u
Z L
AE wdx + axwdx = 0
0 dx 2 0
L du l
Z L
d 2u
Z
du dw
AE
2
wdx = AE w − AE dx ⇒
0 dx dx 0 0 dx dx
Z L Z L
d 2u
du du du dw
AE 2 wdx = AE (L)w(L) − AE (0)w(0) − AE dx
0 dx dx dx 0 dx dx
n
X
u(x) = uj Nj (x)
j=1
w is chosen to be of the same form as the approximate solution (but with arbitrary
coefficients wi ),
n
X
w(x) = wi Ni (x)
i=1
L n n n Z L X n
dNj (x) X dNi (x)
Z X X
AE uj wi dx = R wi Ni (L) + ax wi Ni (x)dx
0 j=1
dx i=1
dx i=1 0 i=1
n Z L Z L
X dNj (x) dNi (x)
AE dx uj = RNi (L) + axNi (x)dx i = 1...n
j=1 0 dx dx 0
which is a system of n equations that can be solved for the unknown coefficients uj .
Institute of Structural Engineering Method of Finite Elements II 46
Axially Loaded Bar Example
The matrix form of the previous system can be expressed as
Z L
dNj (x) dNi (x)
Kij uj = fi where Kij = AE dx
0 dx dx
Z L
and fi = RNi (L) + axNi (x)dx
0
e 1
K11 K12 0
Ktot = 1
K12 1 2
K22 + K11 2
K12 ⇒
2 2
0 K12 K22
1 −1 0
AE
Ktot = −1 2 −1
h
0 −1 1
Institute of Structural Engineering Method of Finite Elements II 47
Axially Loaded Bar Example
After the vectors are formulated we proceed with solving the main equation
Ku = f ⇒ u = K−1 f.
The results are plotted below using 3 elements: